Hi Jonas, On 12-08-2026 12:47, Jonas Smedegaard wrote:
This issue seems flagged as an port-specific ignorable issue, so closing this as not an issue with the package.
I'd expect you to think that a hang is something to worry about on an official supported architecture. But I guess I was wrong.
Nothing has changed in the package and no compile-time warnings or errors have been noticed that caused this change of behavior.
Something changed between 2025-10-20 23:12:59 UTC and 2025-12-06 05:19:56 UTC. Unfortunately the logs have been recycled, so we can't see what the failure was before 2025-10-20 23:12:59 UTC. I'll note that this aligns with a similar change in src:asterisk, which also started to time out, between 2025-10-15 09:11:40 UTC (PASS) and 2025-10-29 14:44:56 UTC.
If you disagree then please reopen,
I agree that having it at serious was a bit exaggerated, sorry for that. But I think it's a bug nevertheless, so I don't understand why you close it without any resolution.
but also I would then appreciate suggestions as to how such an issue might be differently solved. E.g. would you find it better that this package hardcodes that autopkgtest should be skipped for the failing port, or to request ftpmasters drop the package on the port with spuriously hanging autopkgtest?
I don't understand why you call it spurious, could you elaborate? I'd expect an alternative could be to fix the underlying issue? The s390x porters are quite responsive if you need help with s390x specific knowledge.
Paul
OpenPGP_signature.asc
Description: OpenPGP digital signature
